Ad Placement in E-Newspapers Affects Memory, Attitude

Ad Placement in E-Newspapers Affects Memory, Attitude

Rodgers, Shelly

Findings reveal that sponsors may benefit more from advertisements at the beginning or middle of news content, depending on their marketing goals.

Newspapers and advertisers have historically shared a symbiotic relationship where advertising pays for news content and editorial content lends credibility to advertisers who want to reach potential consumers.1 Despite pressure from some advertisers,2 most newspapers are careful about not letting advertisers pretend their messages are news stories.3 However, studies suggest that integrative forms of advertising, such as advertorials or product placements, may confuse readers or erode the credibility of media organizations.4

Internet sponsorships are an integrative form of advertising that are used by many electronic newspapers (e-newspapers).5 Internet sponsors attempt to leverage the e-newspaper’s credibility by placing sponsorships at the top, middle or bottom of a news story. The concern is that sponsorship timing relative to the news story will confuse readers or hurt the e-newspaper’s credibility.6 Although several studies have examined the effects of Internet sponsorships on sponsors, it is also important to examine such effects on enewspapers. If sponsor timing hurts the e-newspaper’s credibility, these findings can inform e-newspapers about advertising policies regarding Internet sponsorships.

The purpose of this research is to examine the effects of sponsor timing on memory for the sponsor and news stories and attitudes toward the sponsor, news stories and e-newspaper. This was accomplished with a 3 (sponsor timing) x 2 (news story) between-subjects experiment using a group of undergraduate students.

Literature Review

Sponsorship refers to the investment in social causes or events that support corporate objectives, such as enhancing corporate reputation or increasing brand awareness.7 Sponsors can associate their name or brand with social causes or events to reach specific target audiences by providing cash and other financial support.8 Sponsorship is fundamentally different from traditional advertising because sponsors indirectly persuade target audiences.9 For instance, traditional advertising seeks to affect consumers directly through the favorable presentation of products. However, sponsorship tries to enhance the perception of the brand and corporate sponsor by associating with events or social causes that are already highly valued among target audiences with the expectation that credibility will transfer from events/causes to the sponsors themselves.10

A 2003 survey by MORI Research showed that 83 percent of e-newspaper users value newspaper Web sites as the best resource for local news, things to do and shopping information, making e-newspapers a desirable outlet for sponsors.” Many sponsors pay to associate with news content in an attempt to leverage the credibility of e-newspapers to enhance their brand. Internet sponsorships are a brief text ad (e.g., “This page is brought to you by HP Computers”), which are easily integrated with news stories.12 Sponsors commonly seek the highest position on the screen, at the top just before the news story, to increase attention to their brand. Some Internet sponsorships are integrated with news content (e.g., in the middle), whereas others are placed at the bottom of the screen, just after the news story in the hope that the sponsorship will be the last thing seen and, therefore, remembered by potential consumers.

Studies that have examined sponsorship effects online have revealed significant benefits for sponsors of content Web sites. One study showed, for example, that sponsors whose products matched news content had higher memory, attitudes and purchase intentions for the sponsored brand than did sponsors/stories that did not match.13 The study further revealed that sponsorship linkage was partially mediated by the e-newspaper’s credibility and intent to return to the e-newspaper, suggesting that sponsors benefited from the site’s credibility.

Another experiment tested the effects of Internet sponsorships on the sponsor and Web site.14 The findings revealed modest benefits to the sponsor and advertiser but few to no benefits to the Web site when a sponsor was present. Specifically, the findings showed more favorable attitudes and higher revisit intentions toward the Web community when there was no sponsor or advertiser, and participants’ revisit intentions were lower for sponsored content versus content that was paid for by a banner ad. The author suggested that sponsorships may be processed differently than other Internet ads and perhaps due to their subtle nature, may confuse readers about the sponsor’s intent, thus resulting in decreased processing of the site’s content.

Although sponsor timing was not examined, we expect that the sponsor’s location relative to the news story will influence individuals’ processing of the sponsors and news content. This expectation is supported with research on source credibility. Starting with Hovland, research on source credibility has examined the effects of source notification on memory and attitude change.15 Research has shown that source timing before or after a message affects memory for the message and source, as well as the persuasive influence of the message.16 Timing of source notification, thus, serves as a cue that audiences use to make sense of the source and message.17

Sponsorship timing makes readers aware of the commercial source for sponsorships to make judgments about the credibility of the source. Research on the effects of advertorial labeling has shown that early presentation of the source or sponsor is needed to increase source credibility and avoid a sleeper effect.18 Cameron argues that sources that are introduced before the message are encoded semantically as part of the message. When a source has higher credibility, the message is perceived as more credible and cognitive processing of the message increases.19 However, in the absence of sponsor or product information, readers may find it difficult to assess the credibility of an Internet sponsor.20 In these instances, the message may be used to assess the credibility of the source.21

Thus, Internet sponsorships positioned prior to news stories may benefit from the perceived credibility of the news content. Due to their brevity and subtlety, however, prior sponsor timing will be integrated at the time the message is encoded and the sponsor may be forgotten. In effect, prior sponsor attribution may confuse readers as to the sponsor’s intention (i.e., did the sponsor influence or provide input into the content?), which may “blur” the line between news and advertising. In such instances, the message will be used to draw conclusions about the sponsor and the sponsor will subsequently benefit from editorial credibility. However, memory for the sponsor may suffer as a result, and the credibility of the news content and e-newspaper will decrease.

In contrast, sponsors in the middle location will interrupt cognitive processing of the news story, laying an additional memory trace for the sponsor resulting in higher memory for the middle timing. However, the interruption will lead to partial integration of the news story, which will result in lower memory for the content and lower credibility for the sponsor and content. In effect, the middle sponsor timing, because of its interruption in the flow of news content, will be the easiest to recognize and interpret. However, the interruption may have an unintended effect that lowers the credibility of the sponsor, content and possibly e-newspaper.

When a source cue is given after a message, encoding processes are already done. A post-source cue is therefore not as memorable as a prior-source cue and will not benefit from the persuasive impact of the message.22 Hence, Internet sponsorships that appear after a news story may not be as memorable as sponsors in the middle timing because of a lack of source/message integration. Ending sponsors will subsequently be stored as episodic rather than semantic memory traces, which are more delicate and, hence, more easily forgotten. Although the benefits to the sponsor may be minimized, the end timing may be the most beneficial for the credibility of the news content and e-newspaper. Without interference of a sponsor, cognitive resources will be devoted to the news story, resulting in greater memory for the content. News stories that are processed without the presence (or influence) of a sponsor may also benefit from having greater perceived credibility. This discussion translates to the following hypotheses:



Memory for the sponsor will be highest for the middle (versus beginning or ending) timing.


Attitude toward the sponsor will be more negative in the middle and more positive in the beginning (versus ending) timings.


Memory for the content will be highest for the ending sponsor (versus beginning or middle) timing.


Credibility of the content and e-newspaper will be the most positive for the ending (versus beginning or middle) timing.


The participants were 114 undergraduates (65 females and 49 males) from a large, Midwestern university. Participants were recruited from journalism and advertising classes and were offered extra credit for their participation. Subjects were treated in accordance with university guidelines involving human subjects.

A 3 (sponsor timing) x 2 (news story) between-subjects experiment was conducted. Sponsor timing was defined as the proximal location of the sponsorship relative to the news story. There were three levels: beginning, middle and end. Beginning sponsor timing occurred below the masthead and before the byline and news story. Middle sponsors were placed in the middle of a news story just after a natural break. End sponsor timing occurred after the last sentence of the news story. News story was a replication variable. Subjects were randomly assigned to one level of sponsor timing and read two news stories that contained the same sponsor/sponsor timing. This increased the power of the design (resulting in a higher N) and the ability to generalize beyond a single news story.23

The experiment took place in a laboratory with 20 computers. Computer settings were standardized prior to the experiment, including screen resolution, color palette, font size and refresh rate. One of the six e-newspapers was randomly selected and loaded onto every computer in the lab. The first group of participants saw the same condition. This decreased the possibility that participants who looked at others’ screens would guess the purpose of the study. After the group finished, another condition was randomly selected and loaded onto the computers. The next group of participants saw that condition. Random assignment continued until nearly equal groups of individuals were exposed to the conditions.

Two news stories were selected after an extensive pre-test.24 Visa credit card was selected from a dozen brands as the sponsor of the news stories after an extensive pre-test. A real (versus fictitious) brand was selected to increase external validity. However, it was important to select a neutral sponsor, because a sponsor with a vested interest in the content could confound the effects of sponsor timing. A pre-test confirmed that Visa was perceived as a neutral sponsor (i.e., did not have a vested interest in the content).

One e-newspaper was created and six versions representing the conditions (3 x 2) were created to control for an ordering effect. The Web site was designed to look like a local e-newspaper in the same Midwestern city. The instructions, news stories and outcome measures were placed online in that order. The sponsorship was a brief line of text, followed by the sponsor’s name (e.g., “News you can use, brought to you by: Visa”). In accordance with ASNE guidelines for advertising labeling in print newspapers, a thin black rule was placed just after the sponsorship.

The e-newspaper was designed with a timekeeper. The timer automatically forwarded individuals to the stories and questionnaire. Stories were timed based on averaging the scores of a slow and fast reader prior to the study. The timer eliminated the need for a stopwatch and enabled participants to start at their leisure.

Dependent Measures

Memory for the sponsor. Memory for the sponsor was measured with two unaided and three aided recall measures. First, individuals were asked to list any brands and/or products they remembered (unaided). Then, individuals selected the correct brand/product out of a list of brands/products (aided recall).25 Scores ranged from 0 to 5, with a higher value indicating higher recall. The average was 1.75 (SD = 1.30).

Memory for the content. Memory for the content was measured with three aided recall items. Participants were asked to select the theme of the story from a list. They were also asked two questions about the content in a multiple choice fashion. Scores ranged from 0 to 3, with a higher value indicating better memory performance. The average was 2.18 (SD = .72).

Attitude toward the sponsor. Attitude toward the sponsor was measured with three, 5-point semantic differential scales. The items: not credible/credible, not believable/believable and not trustworthy/trustworthy (alpha = .88).26 The mean was 9.25 (SD = 3.44).

Attitude toward the content. Attitude toward the content was measured with three, 5-point semantic differential scales. The items: not credible/credible, not believable/believable and not convincing/convincing (alpha = .84).27 The mean was 10.85 (SD = 2.79).

Attitude toward the e-newspaper. Attitude toward the e-newspaper was measured with three, 5-point semantic differential scales. The items: not credible/credible, not believable/believable and not responsible/responsible (alpha = .82).28 The mean was 11.08 (SD = 2.95).


The design of the study required the performance of a Multivariate Analysis of Variance (MANOVA). Sponsor timing was entered as the independent variable and the dependent variables (entered simultaneously) were memory for the sponsor and content and attitude toward the sponsor, content and enewspaper. The overall multivariate test for sponsor timing was statistically significant [F (5, 221)=4.05, p

Tukey’s post hoc comparison test was performed for statistically significant findings to examine differences among the three levels of the independent variable. The analysis revealed that memory for the sponsor significantly differed for the middle/beginning timings (p


Although Internet sponsorships help pay for news content, their effects have been examined primarily in terms of sponsor benefits and not benefits to the e-newspaper. The research presented here attempted to overcome this limitation by offering a conceptual framework that can be used to assess the influence of Internet sponsorships on both the sponsor and e-newspaper. The findings and contribution of this research include the following:

* It confirmed that memory for the sponsor was highest in the middle and lowest in the beginning sponsor timing.

* It empirically demonstrated that the middle timing had a negative effect, whereas the beginning timing had a positive effect on attitude toward the sponsor.

* It showed that memory for and attitude toward the news content were highest for the ending sponsor timing. The effects were marginally significant and the means were in the expected direction.

* It revealed no significant differences for the effects of sponsor timing on the e-newspaper’s credibility.

Theoretical Implications

Research on source credibility asserts that prior-source cues are needed to increase source credibility and to avoid a sleeper effect. Our findings supported this theory to some degree. Attitude toward the sponsor was highest for the beginning timing and lowest for the ending timing. Consistent with source credibility effects, the beginning sponsor was integrated with the news content at the time of encoding and was evaluated with higher credibility. The ending sponsor was processed after encoding of the content had already occurred and was not integrated with the content, and the sponsor’s credibility suffered. Although past studies have shown that individuals will use source cues to make judgments about the source’s credibility, our findings suggest that individuals may also use message cues to make judgments about the source.

Contrary to research on source credibility, memory was lowest for the prior-source cue. We believe that beginning sponsors were processed semantically with the news content and, perhaps because of their small, integrative format, were more easily forgotten than the middle sponsor timing. An implication is that prior-source cue may not always be the best placement for sources and that the nature of the source may interact with source notification to influence memory processes.

Memory was highest for the middle sponsor timing. Why? Unlike the beginning sponsor timing, middle sponsors were integrated with only part of the news content when they appeared half-way through the story. The middle timing essentially interrupted cognitive processing of the content and laid an additional memory trace, which increased sponsor recall. It is unclear from these findings whether the middle source cue was processed semantically or episodically. It is possible that the middle sponsor was processed episodically with the first half of the story, like a post-source cue, but was processed semantically with the remaining half of the story, like a prior-source cue. This finding adds to the literature on source credibility and suggests that memory processes might be heightened by persuasive tactics that work in tandem to influence both semantic and episodic memory processes.

Last, our findings revealed that sponsor timing influenced memory for and attitude toward the content. When sponsors appeared at the end of the news story, more details were recalled and credibility for the content was higher. These findings suggest that sponsor timing may serve as an interruptive cue, which affects elaboration of the news story or sponsor depending on its proximal position relative to the news story. Sponsors that come after the news story are best when more extensive elaboration of the news story is desired.

Practical Implications

The perspective of this research is not only to understand and measure the effects of sponsorships within the context of e-newspapers, but to interpret those findings into solutions for real-world problems. By examining potential benefits to the sponsor and e-newspaper, this investigation opens a new realm of viewing and managing Internet sponsorships. Our findings suggest that there are benefits for advertisers that sponsor news content, but with several drawbacks to the e-newspapers.

First, sponsors placed before the news story will benefit from the credibility of the e-newspaper; whereas, sponsors placed in the middle of the news story will enjoy greater brand awareness. However, the findings suggest trade-offs with these two locations. Whereas beginning sponsors have higher credibility, they are also more easily forgotten. Although middle sponsors are remembered more easily, they have lower credibility. An implication is that advertisers will need to select the sponsor location that meets their marketing goals. When brand awareness is the goal, advertisers should select the middle location. When image enhancement is a goal, the beginning location may be the best choice.

The findings also have implications for the news industry. The concern is that sponsor timing will confuse readers about whether the sponsor had input into the news story, thus decreasing trust in the e-newspaper. Although enewspaper credibility was not significantly affected by sponsor timing, the fact that beginning sponsors were remembered less than middle sponsors may indicate a blurring effect (i.e., beginning sponsors could not be distinguished from news content). Content was remembered and trusted most for sponsors that appeared after the news story, indicating that ending sponsors were perhaps the least harmful to the news.

Rather than forbidding all sponsorships, our findings imply that e-newspaper managers should weigh potential benefits and drawbacks for both the sponsor and e-newspaper and institute advertising policies to reflect this. On the newspaper side, our findings suggest that sponsors that interrupt or come before the story may have to be deciphered by a reader, which leaves more room for misinterpretation. Current advertising policies can be updated to reflect this, or, alternatively, new policies about appropriate forms of sponsorships can be created to ensure that readers are clear about where advertising ends and editorial begins.

Limitations and Directions for Future Research

Despite the contributions of this research, a number of limitations should be noted. First, participants were a convenience sample of students who may not be representative of the larger population of adults. Future studies can use different student and non-student samples to determine the robustness of our findings. Second, although an extensive pre-test led to the selection of a well-known brand, the established brand may have heightened attitudes toward the sponsor, thereby suppressing attitudes toward the content and e-newspaper. This could account for the fact that sponsor timing did not significantly influence the e-newspaper’s credibility. Fortunately, these limitations can be overcome by using different established and fictitious brands in follow-up studies.


Although the literature primarily addresses sponsorship effects from the advertiser’s perspective, proper identification of sponsors is also a key concern for the news industry. It is clear from these findings that sponsor timing can benefit sponsors but with some drawbacks to the e-newspaper. The findings imply that advertisers and newspapers must compromise so that advertisers can attain their marketing goals without exploiting the e-newspaper and its contents.


1. Advertising and the World Wide Web, eds. David W. Schumann and Esther Thorson (Hillsdale, NJ: Lawrence Erlbaum Associates, 1999) 159-173.

2. Lawrence C. Soley and Robert L. Craig, “Advertising Pressures on Newspapers: A Survey,” Journal of Advertising 21, no. 4 (winter 1992): 1-9.

3. Kathleen T. Lacherand Herbert J. Rotfeld, “Newspaper Policies on the Potential Merging of Advertising and News Content,” Journal of Public Policy & Marketing 13, no. 2 (fall 1994): 281-289.

4. Glen T. Cameron and Patricia A. Curtin, “Tracing Sources of Information Pollution: A Survey and Experimental Test of Print Media’s Labeling Policy for Feature Advertising,” Journalism & Mass Communication Quarterly 72, no. 1 (spring 1995): 178-189; Dennis M. Sandier and Eugene Secunda, “Point of View: Blurred Boundaries-Where Does Editorial End and Advertising Begin,” Journal of Advertising Research 33, no. 3 (May/June 1993): 73-80. For a discussion, see also Wendy S. Williams, “The Blurring of the Line Between Advertising and Journalism in the On-Line Environment,” in The Electronic Grapevine: Rumor, Reputation, and Reporting in the New On-Line Environment, eds. Diane L. Borden and Kerrie Harvey (Mahwah, NJ: Lawrence Erlbaum Associates, 1998) 31-41.

5. Brill, “Online Newspaper Advertising”; Shelly Rodgers, “The Effects of Sponsor Relevance on Consumer Reactions to Internet Sponsorships,” Journal of Advertising 32, no. 4 (winter 2003): 67-76.

6. Katherine Yung, “Sponsored Web Sites Raises Questions of Ethics,” St. Paul Pioneer Press, Tech Industry Section (January 4, 1999); Martha L. Stone, “The Wall Crumbles: The Defined Line Between Print and Advertising is Gone,” Editor & Publisher, February 1999, p. 20.

7. Dwane Hal Dean, “Associating the Corporation with a Charitable Event Through Sponsorship: Measuring the Effects on Corporate Community Relations,” Journal of Advertising 31, no. 4 (winter 2002): 77-87; Meryl P. Gardner and Philip Shuman, “Sponsorships and Small Businesses,” Journal of Small Business Management 26, no. 4 (October 1988): 44-52.

8. T. Bettina Cornwell and Isabelle Maignan, “An International Review of Sponsorship Research,” Journal of Advertising 27, no. 1 (Spring 1998): 1-21.

9. James Crimmins and Martin Horn, “Sponsorship: From Management Ego Trip to Marketing success,” Journal of Advertising Research 36, no. 4 (July/August 1996): 11-21.

10. Crimmins and Horn, “Sponsorship”; Dwane Hal Dean, “Associating the Corporation with a Charitable Event Through Sponsorship: Measuring the Effects on Corporate Community Relations,” Journal of Advertising 31, no. 4 (winter 2002): 77-87; Kevin P. Gwinner and John Eaton, “Building Brand Image Through Event Sponsorship: The Role of Image Transfer,” Journal of Advertising 28, no. 4 (winter 1999): 47-57; Stephen R. McDaniels, “An Investigation of Match-Up Effects in Sport Sponsorship Advertising: The Implications of Consumer Advertising Schemas,” Psychology & Marketing 16, no. 2 (March 1999): 163-184.

11. For additional evidence, see Jennifer Greer and Donica Mensing, “U.S. News Web Sites Better, But Small Papers Still Lag,” Newspaper Research Journal 25, no. 2 (spring 2004): 98-112.

12. Shelly Rodgers and Esther Thorson, “The Interactive Advertising Model: How People Perceive and Process Interactive Ads,” Journal of Interactive Advertising 1, no 1 (fall 2000), (5 November 2004).

13. Rodgers, “The Effects of Sponsor Relevance.”

14. Karen L. Becker-Olsen, “And Now, A Word From Our Sponsor: A Look at the Effects of Sponsored Content and Banner Advertising,” Journal of Advertising 32, no. 2 (summer 2003): 17-32.

15. Carl I. Hovland, Irving L. Janis, and Harold H. Kelly, Communication and Persuasion (New Haven: Yale University Press 1953).

16. Pamela M. Homer and Lynn R. Kahle, “Source Expertise, Time of Source Identification, and Involvement in Persuasion: An Elaborative Processing Perspective,” Journal of Advertising 19, no. 1 (spring 1990): 30-39; Chester A. Insko, “Primacy Versus Recency in Persuasion as a Function of Timing of Arguments and Measures,” Journal of Abnormal and Social Psychology 63, (1964): 381-391; Anthony R. Pratkanis, Anthony G. Greenwald, Michael R. Leippe, and Michael H. Baumgardner, “In Search of Reliable Effects: III. The Sleeper Effect is Dead. Long Live the Sleeper Effect,” Journal of Personality and Social Psychology 54, (1988): 203-218.

17. Michael D. Slater, Donna Rouner, David Karan, Kevin Murphy, and Frederick Beauvais, “Placing Alcohol Warnings Before, During, and After TV Beer Ads: Effects on Knowledge and Responses to the Ads and the Warnings,” Journalism & Mass Communication Quarterly 76, no.3 (autumn 1999): 468-484.

18. Glen T. Cameron, “Does Publicity Outperform Advertising? An Experimental Test of Third Party Endorsement,” Journal of Public Relations Research 6, no.3 (1994): 185-207.

19. Cameron, “Does Publicity Outperform Advertising?”

20. Gita Venkataramani Johar and Michael Tuan Pham, “Relatedness, Prominence, and Constructive Sponsor Identification,” Journal of Marketing Research 36, no. 3 (August 1999): 299-312. For instance, the findings from this study suggest that proper sponsor identification involves a substantial degree of construction.

21. Slater et al., “Placing Alcohol Warnings.”

22. Cameron, “Does Publicity Outperform Advertising?”

23. William D. Wells, “The Perils of N=1,” Journal of Consumer Research 28 (December 2001 ):494-498.

24. A total of 40 news stories were initially collected from 20 of the largest U.S. daily e-newspapers. The stories were pre-tested on key items (e.g., credibility, relevance, likeability) with a convenience sample of undergraduates. Two stories with the highest means were selected. One editor and two journalists edited the final stories to control for length and tone. A final pro-test revealed no significant differences between the stories in terms of the key items.

25. See Cameron and Curtin, “Tracing Sources of Information Pollution.”

26. Eric Haley, “Exploring the Construct of Organization as Source: Consumers’ Understandings of Organizational Sponsorship of Advocacy Advertising,” Journal of Advertising 25, no.2 (summer 1996): 19-35.

27. Cecilie Gaziano and Kristin McGrath, “Measuring the Concept of Credibility,” Journalism Quarterly 63, no. 3 (autumn 1986): 451-462.

28. Gaziano and McGrath, “Measuring the Concept of Credibility.”

Rodgers is an assistant professor and Cameron is a professor and holds the Maxine Gregory Chair. Both are in the School of Journalism at the University of Missouri.

Brill is an associate professor and the interim dean in the William Allen White School of Journalism and Mass Communication at the University of Kansas.

Copyright Newspaper Research Journal, Department of Journalism, University of Memphis Winter 2005

Provided by ProQuest Information and Learning Company. All rights Reserved